nulla-hive-mind is a local-first AI app for research and shared knowledge. It keeps your work on your own computer first. It can help with research tasks, remember past work, and share knowledge with other devices on your network.
You can use it to:
- Ask AI questions on your PC
- Run research tasks in the background
- Keep notes and memory between sessions
- Connect with other nodes on a network
- Use local AI models or API-compatible models
- Work with privacy in mind
For Windows, use:
- Windows 10 or Windows 11
- At least 8 GB of RAM
- 10 GB of free disk space
- A stable internet connection for the first download
- A recent GPU helps with local AI, but it is not required
For best results:
- 16 GB of RAM or more
- An SSD
- A modern CPU with 4 or more cores
- Open the release page
- Find the latest Windows file
- Download the file to your computer
- If the file is a
.zip, right-click it and choose Extract All - Open the extracted folder
- Double-click the app file to run it
- If Windows asks for permission, select Yes
If you see more than one file, pick the Windows version that matches your computer. For most users, that is the standard 64-bit build.
When you open nulla-hive-mind for the first time:
- Let it finish setting up its local files
- Choose a model source if asked
- Keep the default settings if you are not sure
- Wait for the main screen to load
The first launch may take a little longer because the app prepares its local memory and tools.
After the app opens, check these settings:
- AI model: Use a local model or an OpenAI-compatible API
- Search source: Connect to SearxNG if you want web search
- Memory: Leave persistent memory on if you want the app to remember past work
- Sandbox: Keep sandboxed execution on for safer task runs
- Network: Turn on peer-to-peer options if you want other nodes to find each other
If you are using a local model:
- Install Ollama first if the app asks for it
- Make sure the model is available on your machine
- Pick a smaller model first if your PC is older
Your data stays on your device first. This helps when you want more control over your files and research.
The app can store useful context across sessions. That helps it remember long tasks, notes, and past findings.
You can give it a task, then let it collect and organize information for you.
The app can share data with other nodes. This helps build a knowledge mesh across systems you trust.
The app can run tasks in a safer space. This helps limit what a task can do on your computer.
You can connect tools that speak the same API style as OpenAI-compatible services.
Try these steps:
- Right-click the file and choose Run as administrator
- Make sure Windows SmartScreen did not block it
- Check that you downloaded the latest release file
- Reboot your PC and try again
Try:
- Extracting the files again
- Running it from a normal folder like
DownloadsorDesktop - Checking that your antivirus did not remove a file
Check:
- Your internet connection
- Your SearxNG settings
- The search server address if you changed it
Try:
- A smaller model
- Closing other heavy apps
- Using a machine with more RAM
- Using GPU support if your setup has it
nulla-hive-mind can work with different tools and services. Common setups include:
- Ollama for local models
- OpenAI-compatible APIs for cloud or self-hosted models
- SearxNG for search
- Peer nodes for shared knowledge
- Local storage for memory and task history
If the app asks for a server address, copy it carefully. A small typo can stop the connection.
A simple flow looks like this:
- Open the app
- Start a new research task
- Give it a clear question or goal
- Let it gather and sort information
- Review the result
- Save useful notes to memory
Good task examples:
- Find the main points of a topic
- Compare a few sources
- Keep notes from a long research session
- Store facts for later use
- Share results with another trusted node
The app may keep data in places like:
- App data folders in Windows
- Local cache folders
- A project folder you choose during setup
If you want to move the app to another PC:
- Copy the app files
- Copy the local data folder if you want to keep memory
- Open the app on the new machine and check the settings
After setup, try this:
- Ask the app to research a simple topic
- Save one result to memory
- Restart the app
- Check that the saved item still appears
This helps confirm that the core setup works.
The app is built for local-first use. That means you can keep more data on your own machine and reduce what you send to outside services.
To stay in control:
- Use local models when you can
- Review network settings before sharing data
- Keep peer-to-peer sharing to trusted systems
- Save only the memory you need
If something still does not work, check:
- You downloaded the latest release
- You used the correct Windows file
- The files are fully extracted
- Windows did not block the app
- Your antivirus did not quarantine a file
- Your model path or API key is correct
- Your search service is online
- You have enough free disk space
Get the Windows build here and download and run this file from the latest release page:
agent-runtime, ai-agents, autonomous-research, decentralized-ai, dht, distributed-systems, knowledge-mesh, multi-agent-systems, nat-traversal, ollama, openai-compatible, openclaw, peer-to-peer, persistent-memory, privacy-first, privacy-first-ai, python, sandboxed-execution, searxng